* Added modeld.py (WIP)
* No more VisionIpcBufExtra
* Started work on cython bindings for runmodel
* Got ONNXModel cython bindings mostly working, added ModelFrame bindings
* Got modeld main loop running without model eval
* Move everything into ModelState
* Doesn't crash!
* Moved ModelState into modeld.py
* Added driving_pyx
* Added cython bindings for message generation
* Moved CLContext definition to visionipc.pxd
* *facepalm*
* Move cl_pyx into commonmodel_pyx
* Split out ONNXModel into a subclass of RunModel
* Added snpemodel/thneedmodel bindings
* Removed modeld.cc
* Fixed scons for macOS
* Fixed sconscript
* Added flag for thneedmodel
* paths are now relative to openpilot root dir
* Set cl kernel paths in SConscript
* Set LD_PRELOAD=libthneed.so to fix ioctl interception
* Run from root dir
* A few more fixes
* A few more minor fixes
* Use C update_calibration for now to exactly match refs
* Add nav_instructions input
* Link driving_pyx.pyx with transformations
* Checked python FirstOrderFilter against C++ FirstOrderFilter
* Set process name to fix test_onroad
* Revert changes to onnxmodel.cc
* Fixed bad onnx_runner.py path in onnxmodel.cc
* Import all constants from driving.h
* logging -> cloudlog
* pylint import-error suppressions no longer needed?
* Loop in SConscript
* Added parens
* Bump modeld cpu usage in test_onroad
* Get rid of use_nav
* use config_realtime_process
* error message from ioctl sniffer was messing up pyenv
* cast distance_idx to int
* Removed cloudlog.infos in model.run
* Fixed rebase conflicts
* Clean up driving.pxd/pyx
* Fixed linter error
old-commit-hash: 72a3c987c0
* Clean up model_publish args to simplify cython bindings
* pass by reference
* Move FCW and model confidence queues into PublishState
old-commit-hash: 663fc0d8fe
* Fix flicker in nav enabled state
* Move all relevant information for validity checks into navModel packet
* Ignore locationMonoTime in replay tests
* Check route valid in navmodeld
* sm update
* check that
* update refs
---------
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: e346839c68
* Started work on model runner refactor
* Fixed some compile errors
* everything compiles
* Fixed bug in SNPEModel
* updateInput -> setInputBuffer
* I understand nothing
* whoops lol
* use std::string instead of char*
* Move common logic into RunModel
* formatting fix
old-commit-hash: c9f00678af
* 0.7
* magic
* faster magic
* more simple
* up
* empty
* more mid bits
* naive
* flatten
* dz
* that can stay
* this is fine
* what the
* what the
* giRevert "what the"
This reverts commit 1619ba68e6.
* Revert "what the"
This reverts commit 0037dd3682.
* 1x fine
* that was fine
* combined
* independent cum
* 0 is fine
* use metrics
* up cereal
* process and publish from modeld
* cleanup
* use s.output
* bg
* a greener approach
* dns
* serial
* update ref commit
* rebase
* ref
* cereal master
---------
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: ce2dffe566
* 10s lat
* Full length MPC
* redfine N
* Leave controls the same for now
* Updates
* use long plan in lat plan
* interp plan
* simplergit add selfdrive/controls/plannerd.py selfdrive/controls/
* expand to 10s
* revert this
* fix linter
* vizualize
* fix long test
* typo
* cleanup
* compiles
* unused
* unused
* bump cereal
* bump cereal
* use model if no uiplanm
* update replay
* update ref commit
* bump cereal to master
old-commit-hash: eb8bdc0026
* Added navmodeld
* New nav model: 7c306685-5476-4bd4-ab65-105b01b6bca8/300, feats only
* little cleanup
* Remove NAV flag
* Moved to_kj_array_ptr to commonmodel.h
* Switch from decimation to last_frame_id check
* add to release files
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: bb8a38a050
* refactor
* Add planer modes to support offline, acc, and blended
* add acceleration
* Fix index
* Update model ref
* Read in model outputs
* Add model msg
* Add e2e logic
* Add source
old-commit-hash: e1b7a37a1f
* pc thneed prereqs
* ugh, out of date
* that can stay private
* memcpy here is fine in SNPE variant
* release files
* thneed docs don't work anymore. they didn't look too useful
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: b6e355a933
* driverview
* auto choose
* useless
* remove
* modeld not use toggle
* remove from params
* should have commissue anyway
* delete icon file
* deal with init more elegantly and prohibit switching when engaged
* nonblock save when not unstable
* filter very low speeds
* a few more secs dont harm
* draw icon
* match
* radius
* use const
* ref commit
* new transs
* put_bool_nonblocking
* 5k route tune
* ref comm
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 22408781f7
* rgb to nv12
* nv12 works (w memcpy)
* correct now
* no copy
* fix nv12 with fast debayer
* reverts of unused stuff
* ui use nv12
* comment out thumbnails for now
* rebase fix
* dm read nv12
* model read nv12
* fix ffmpeg encoder
* thumbnails from nv12
* replay to nv12
* python framereader support nv12
* remove hardcoded frame/buffer sizes
* fix build
* ffmpeg encoder fix buffers
* small cleanup
* reduce power usage test
* fix cpu usage test
* fix snapshot
* fix loggerd test
* bump cereal
Co-authored-by: Comma Device <device@comma.ai>
Co-authored-by: Joost Wooning <jwooning@gmail.com>
old-commit-hash: ea5b8cdfb1
* Do not run model when skipping frames
* also prepare wide
* add logging
* remove newline
* small skip test
* more random
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 3741908733
* Added wide cam vipc client and bigmodel transform logic
* Added wide_frame to ModelState, should still work normally
* Refactored image input into addImage method, should still work normally
* Updated thneed/compile.cc
* Bigmodel, untested: 44f83118-b375-4d4c-ae12-2017124f0cf4/200
* Have to initialize extra buffer in SNPEModel
* Default paramater value in the wrong place I think
* Move USE_EXTRA to SConscript
* New model: 6c34d59a-acc3-4877-84bd-904c10745ba6/250
* move use extra check to runtime, not on C2
* this is always true
* more C2 checks
* log if frames are out of sync
* more logging on no frame
* store in pointer
* print sof
* add sync logic
* log based on sof difference as well
* keep both models
* less assumptions
* define above thneed
* typo
* simplify
* no need for second client is main is already wide
* more comments update
* no optional reference
* more logging to debug lags
* add to release files
* both defines
* New model: 6831a77f-2574-4bfb-8077-79b0972a2771/950
* Path offset no longer relevant
* Remove duplicate execute
* Moved bigmodel back to big_supercombo.dlc
* add wide vipc stream
* Tici must be tici
* Needs state too
* add wide cam support to model replay
* handle syncing better
* ugh, c2
* print that
* handle ecam lag
* skip first one
* so close
* update refs
Co-authored-by: mitchellgoffpc <mitchellgoffpc@gmail.com>
Co-authored-by: Harald Schafer <harald.the.engineer@gmail.com>
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 85efde269d
* model output struct for metadata
* better max element search
* ModelDataRawDesireProb.to_array()
* eliminate some copies
* not worth messing with softmax now
* remove unused includes
* more cleanup
* no longer pointers
* better variable name
* fix recurrent state
* improve variable name
* fix OUTPUT_SIZE and NET_OUTPUT_SIZE
old-commit-hash: 40d2e4ec90
* convert model data for leads to struct
* make things more consistent
* change names with path to plan
* consistent use of plural
old-commit-hash: 7bb13acc45
* lane lines and road edges
* roll back some changes to find what broke things
* rollback more changes to find issue
* fix order of lane line probs
* try outputs on stack again
* initialize output array
old-commit-hash: 755a0a63a2
* model pose more struct
* constexpr seems useful
* more concise
* fix order
* seems more readable
* plan struct
* attempt to fix compiling on device
* fix constexpr errors on device
* fix rotation rate in log
* maybe this fixes compiling constexpr on device
* exp() cannot be used in constexpr on c2
* only copy what we actually log
* simplify pivot()
* fix setting position std
* see if this works
* try lambda again
* giving up on the lambda to select member
* fix position std
* fix param order
old-commit-hash: a7f36c9daf
* 6be443f2-ed70-4580-9c31-61b94d068e24/950
* meta changed too
* new ref
* new ref from CI
* unused
* Revert "new ref from CI"
This reverts commit 28b6bdc810.
* new ref from CI again
old-commit-hash: 72a736f90e
* completely untested
* it builds now
* bug fixes, save 1ms
* using a kernel to copy works
* more sane API to loadyuv
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 83ff9ca331
* use the actual file name
* use macro
* fix typo
* continue
* Apply suggestions from code review
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 7db04ab0c5
* Revert "New desire model (#21729)"
This reverts commit 709932b3bfb4c3b4f33711b8d07615469e762832.
* Updated process replay ref and reverted release notes
old-commit-hash: 15d6bbb47c
* New model: b822ee32-917c-4639-99f8-25be3e6aabad/900
* Enable keeps
* use new packet
* Send keep pulses once per second
* Updated model_replay_ref_commit
* New model: 0f160bf9-1bce-43e6-9785-564b2453eeac
* Bump model_replay_ref_commit
* Bump submodules
* New model: fd904bf9-5dd8-4619-8997-c5668400e078/550
* New model: fd904bf9-5dd8-4619-8997-c5668400e078/950
* No keeps
* Bumped model replay ref
Co-authored-by: Harald Schafer <harald.the.engineer@gmail.com>
old-commit-hash: e83973b9b3
Mitchell Goff
Dean Lee
ZwX1616
YassineYousfi
Harald Schäfer
George Hotz
Adeeb Shihadeh
Willem Melching
Lukas Petersson
Greg Hogan